Currently, computing devices are prone to error when a user provides input to initiate an application of the device because the input is often indeterminate to the device. Furthermore, when providing an audible input sequence such as a command to a device, a user sometimes becomes distracted and, while the user may still be speaking, the user is not intentionally providing input to the device but e.g. speaking to another person. However, the device continues to detect and process audio from the user when the user is speaking to the other person and not to the device, leading to unintentional input to the device and even e.g. execution of undesired commands. In such instances, a user may e.g. be required to provide the desired input to the device again (including e.g. previously input portions) in a new sequence, which can be burdensome.